Background. Schizophrenic patients are known to have neuropsychological deficits including impaired verbal fluency, but it is not clear whether this latter deficit is: (a) a consequence of overall intellectual deficit; (b) shared with affective psychotic patients; or (c) shared by the relatives of schizophrenic patients; and (d) shared by the relatives of affective psychotic patients.

Methods. We administered Thurstone's Verbal Fluency Test to 45 schizophrenic patients and 72 of their relatives, and 30 affective psychotic patients and 53 of their relatives. Subjects were asked to generate as many words as possible beginning with the letters ‘C’ and ‘S’ and the total was taken as the dependent variable. Subjects also completed the National Adult Reading Test (NART) to provide a measure of (pre-morbid) IQ.

Results. Schizophrenic patients generated significantly fewer words than affective psychotic patients, however adjusting for NART this became non-significant. Schizophrenic (but not affective psychotic) patients generated significantly fewer words than their relatives; again adjusting for NART this became non-significant. Patients who had been exposed to obstetric complications (OC+) and those who had not (OC−) had similarly poor verbal fluency scores. Relatives of OC+ schizophrenic patients had superior verbal fluency than relatives of OC− schizophrenic patients and this remained significant after adjustment for NART.

Conclusions. The results suggest that some families transmit impairment in verbal fluency as part of a pattern of lower overall IQ. However, in other families, relatives show largely normal neuropsychological function, and the poorer verbal performance of the schizophrenic member appears to have arisen secondary to his/her exposure to OCs.